A Laboratory Diet-Overlay Bioassay to Monitor Resistance in Lygus lineolaris (Hemiptera: Miridae) to Insecticides Commonly Used in the Mississippi Delta

Abstract

A laboratory, diet-overlay pesticide bioassay was developed using a susceptible population of the tarnished plant bug, Lygus lineolaris (Palisot de Beauvois), to study its susceptibility to neonicotinoid, sulfoxamine, organophosphate, and pyrethroid insecticides (thiamethoxam, sulfoxaflor, acephate, and permethrin, respectively). The diet-overlay bioassay was compared to the traditional glass-vial surface residue bioassay. We measured LC₅₀ values by feeding tarnished plant bug adults known doses of insecticides dispensed on top of diet in a 10% solution of honey water for thiamethoxam and 10% acetone in water solutions for permethrin, acephate, and sulfoxaflor. Both the diet-overlay and glass-vial bioassays used dose-response (mortality) regression lines to calculate LC₅₀ values for each insecticide at 6-, 24-, 48-, and 72-h post-exposure. Data variability from the glass-vial bioassay was higher for permethrin, sulfoxaflor, and thiamethoxam than the diet-overlay bioassay, for all evaluation times. In contrast, there was lower variability among replicates to acephate in the glass-vial assay compared to the diet-overlay assay. Control mortalities observed on diet-overlay bioassay were lower (0–5%) than those observed on the glass-vial bioassay (4–27%). The use of green beans, floral-foam, rolling glass vials, and insect handling made the existing standard method tedious to manipulate and difficult to handle large numbers of individuals. The nonautoclaved solid diet provides an opportunity to significantly reduce cost and variability associated with procedures of other bioassay methods. In general, the baseline data provide a basis for future comparison to determine changes in resistance over time.

The tarnished plant bug, Lygus lineolaris (Palisot de Beauvois), is a highly polyphagous pest that feeds on more than 400 plant species (Snodgrass et al. 1984a, 1984b, Young 1986, Parys and Snodgrass 2014). This insect is considered one of the most important pests of cotton, Gossypium hirsutum (L.) (Malvales: Malvaceae), throughout the southern and southeastern U.S. cotton belt (Cleveland 1985; Snodgrass 1996a; Snodgrass and Scott 1999, 2000; Allen et al. 2012; Siebert et al. 2012; Portilla et al. 2014, 2018). Adults and nymphs can injure cotton at any phenological growth stage (Tugwell et al. 1976, Siebert et al. 2012). Cook and Threet (2020) mentioned that during the 2019 growing season, the Lygus complex caused loss of cotton than any other insect species that year, infesting >2.5 million ha of cotton during this year, resulting in a yield loss of >335.000 bales ($164 million). A total of 1.2 million ha of cotton were treated with insecticides in 2018, which increased to almost 1.5 million in 2019. Fleming et al (2016) noted that the number of sprays and the total cost for tarnished plant bug in the Mississippi Delta have dramatically increased over the last decade. The cost of these control strategies has exceeded 10-fold, from $5 million to >$50 million in a time period of 15 yr (Portilla et al. 2018).

Insecticides remain the primary tarnished plant bug control strategy in many southern states. Of particular concern are tarnished plant bug infestations in Tennessee, Missouri, Arkansas, Louisiana, and Mississippi that required from 1.4 to 4.9 insecticide applications per acre, respectively, during the 2008 growing season (Williams 2009). Fleming et al. (2016) found that since 2008 the average number of applications per year has increased to five or more. With this level of treatment, tarnished plant bug has been controlled in most areas, but in some areas the development of insecticide-resistant populations leads to control failures. The increasing level of resistance of tarnished plant bug to common broad-spectrum insecticides in several groups (neonicotinoid, sulfoxamine, organophosphate, and pyrethroid) is well documented. Cleveland and Furr (1980) and Cleveland (1985) found increased resistance to some organophosphates in tarnished plant bug collected in the Mississippi Delta. A continuous survey of pesticide susceptibility of tarnished plant bug populations in the Delta region of the Mid-South began in 1993 (Snodgrass 1994) and has continued to the present (Snodgrass and Elzen 1995; Snodgrass 1996a, 1996b; Snodgrass and Scott 1999, 2000, 2002; Snodgrass et al. 2008a, 2008b, 2009; Allen et al. 2012; Parys et al. 2015, 2017, 2018; Portilla et al. 2018). Since the first report of resistance of tarnished plant bug populations to pyrethroids in 1993 (Snodgrass 1996b) and the cross-resistance of these populations to some organophosphates (Snodgrass and Elzen 1995), the number of tarnished plant bug populations resistant to insecticides have increased and become widespread across Arkansas, Louisiana, and Mississippi. Compiling data to determine the prevalence of resistance for those populations became a problem due to requirements of regular, large-scale monitoring of field population to determine the geographic distribution of resistance.

Insecticide resistance in plant bugs has been studied with topical application of insecticides (Cleveland and Furr 1980, Cleveland 1985) and several different glass-vial bioassays (Brindley et al. 1982; Snodgrass 1996a; Snodgrass and Scott 2000, 2002). Both bioassays rely on the use of green beans, broccoli, and wet floral-foam as a source of food and water. Allen et al. (2012) and Parys et al. (2015) used a diet-incorporated bioassay to assess susceptibility of tarnished plant bug nymphs. Snodgrass et al. (2008a) mentioned that control mortality is an important factor that could affect the accuracy of the LC₅₀ estimates. Portilla et al. (2014) developed a bioassay to evaluate the potential of the insect growth regulator novaluron against tarnished plant bug using a nonautoclaved solid artificial diet dispensing the insecticide over the top of the insect diet (diet-overlay essay). Portilla et al. (2014) reported 0% mortality for adults in the controls after 5 d of exposure. We used this same solid diet assay to compare glass-vial versus diet-overlay assay assessments of the efficacy of the most common neonicotinoid, sulfoxamine, organophosphate, and pyrethroid insecticides (thiamethoxam, sulfoxaflor, acephate, and permethrin, respectively) used in the Mississippi Delta to control tarnished plant bug.

Materials and Methods

Preparation of the Diets

The nonautoclaved, semisolid artificial diet (used for insect colony maintenance) and the solid diet (used for bioassay) were prepared according to procedures described by Portilla et al. (2011) and Portilla et al. (2014), respectively (Supp Appendix 1 [online only]). The prepared semisolid diet was poured into 500 cc plastic dispensing bottles and packed in parafilm bags (8 × 8 cm) (Parafilm Sigma P7793, www.sigmaaldrich.com). To reduce the thickness and enhance the probability for bug feeding, each diet package was physically elongated by stretching the parafilm walls while the diet was warm. The stretched packs were then stored at 15°C for later use. The solid diet was prepared by pouring the final mixture into individual 37-ml plastic cups, adding 5 ml of diet per cup (T-125, Solo Cup Company, www.solocup.com). Diet-filled cups were then held at room temperature to cool and solidify before use.

Insects

All studies were conducted at the USDA-ARS Southern Insect Management Research Unit, Stoneville, MS, in an environmental room with a temperature of 25 ± 2°C, 65 ± 10% RH, and a 12:12 (L:D) h photoperiod. Adults were from a colony established in 1998 (Cohen 2000). The colony was maintained on nonautoclaved semisolid artificial diet following a protocol described by Portilla et al. (2011). To obtain individuals from each generation of similar standards and ages, egg masses were collected in gel oviposition packs placed on top of the screened top rearing containers that held about 2,000 same-aged 6-d-old adults. The oviposition packs were prepared by pouring 50 ml of 4.5% Gelcarin solution (Carrageenan GP 812, www.phytotechlab.com) into a parafilm bag (8 × 8 cm) (Parafilm Sigma P7793, www.sigmaaldrich.com) same bags used for diet packing. Eggs obtained within the first 48 h of the oviposition period were collected in three gel packs and returned to the rearing colony. Then, a second egg collection was obtained using three gel packs within a 24-h time period and used for the bioassays. Seven days after oviposition and before emergence, the gel packs with the collected eggs were placed on shredded paper (to prevent cannibalism) with a stretched diet pack (for feeding as soon nymphs emerge from the gel) inside rectangular storage cages (32 × 25 cm) (Rubbermaid, www.rubbermaid.com), with 25 × 23 cm openings cut into the bases and covered with bleached muslin fabric (Roclon 120″ Pro ID 87264) (www.theonlinefabricstore.com). The top was also ventilated with a 23 × 30 cm opening covered with a removable piece of bleached muslin fabric (45 × 35 cm) for first–fourth instar nymphs. This material was replaced with 1.0 mm mesh netting fabric (mosquito mesh 55″/56″ 0100) (www.theonlinefabricstore.com) to hold fifth instar nymphs and adults. Two stretched diet packs were placed on top of the cages and changed every other day from the day of egg hatch until adults were 2-d-old and ready for use in bioassay. Each rearing container held approximately 2,000 (50F:50M) insects, and from these 1,440 mixed sex adults were selected for running both bioassays simultaneously using the same generation of insects (720 insects for glass vial and 720 insects for diet overlay). Eleven generations were reared for this study. This experiment was repeated 11 times (each time with a different generation of insects). Each generation was considered a replicate. About 1,980 insects were used for each insecticide totaled across all generations. Mortality was measured 6-, 24-, 48-, and 72-h post-exposure to the treated glass vial with or without floral-foam and diet-overlay cup.

Insecticides Used for Assays

Technical grade insecticides of four chemical groups were used in assays: organophosphate (acephate), pyrethroid (permethrin), neonicotinoid (thiamethoxam), and sulfoxamine (sulfoxaflor), each with purity >95%. Insecticides were purchased from Chem Service (West Chester, PA). Glass-vial contact bioassays with acephate, permethrin, and sulfoxaflor were conducted following the procedures developed by Snodgrass (1996a), and the glass-vial oral bioassays with thiamethoxam were conducted following Teague and Tugwell (1996). Insecticides were stored in a freezer at −20°C. Acephate, permethrin, and sulfoxaflor were dissolved in pesticide-grade acetone (Fisher, Fair Lawn, NJ), and thiamethoxam was dissolved in a 10% honey solution. All insecticides were serially diluted. The test concentrations for acephate and permethrin were 0, 1, 3, 10, 30, and 100 μg per vial and for thiamethoxam and sulfoxaflor were 0, 0.5, 1.5, 5, 15, and 50 μg per vial. The same concentrations were used for the diet-overlay bioassay (contact and feeding); however, all insecticides were dissolved in pesticide-grade acetone before being overlaid on the diet. Data variability between bioassays was based on mortality data subjected to Log Dose Probit analysis generate to estimate the lethal concentration (LC₅₀) and resistance ratio (RR₅₀) among generations.

Glass-Vial Contact Bioassays

To determine contact toxicity of acephate, permethrin, and sulfoxaflor, each of the test concentrations was dissolved in acetone and then pipetted in 0.5 ml aliquots into 20 ml scintillation glass vials (2.8 cm diameter, 6.1 cm high). Treated glass vials were then rolled on a hotdog roller (Star MFG, Smithville, TN) without heat under a fume-hood until no more droplets were seen on the glass wall, which resulted in an evenly coated layer of insecticide on each vial. A 4 cm piece of green bean, Phaseolus vulgaris L. (Fabales: Fabaceae), was then washed and sterilized with 10% sodium hypochlorite solution and added to the vial as the food source for the test bugs. One 2-d-old tarnished plant bug adult (sex unknown) was added per vial and thus exposed to one of the five concentrations of each insecticide. Control vials were treated with acetone alone. Each vial was plugged with cotton to confine the insect but allow air circulation. Thirty 2-d-old tarnished plant bug adults (mixed sex) were tested per concentration for each insecticide.

Glass-Vial Feeding Bioassays

Pieces of floral-foam (12 cm diameter) were placed individually in 20 ml glass scintillation vial as described by Snodgrass et al. (2008b). To determine the toxicity of thiamethoxam via feeding, the insecticide was dissolved in a 10% honey solution, which was pipetted in 0.5 ml aliquots onto the floral-foam. One 2-d-old tarnished plant bug adults of unknown sex was placed in each of the test vials, which contained one of the five concentrations of the insecticide. Control vials were treated with 10% honey water solution only. Each glass vial was plugged with cotton to confine the bug but allow air circulation. Thirty 2-d-old tarnished plant bug adults (mixed sex) were tested per each concentration.

Diet-Overlay Contact/Feeding Bioassays

To determine toxicity of acephate, permethrin, thiamethoxam, and sulfoxaflor through contact and feeding on diet, the test insecticides were dissolved in acetone and pipetted the solution in 0.5 ml aliquots into 37 ml SOLO cups (5 ml of diet per cup). Treated diet cups were left under a fume-hood until the residues were dry on the surface of the diet. One 2-d-old tarnished plant bug adult (unknown sex) was added per diet cup on one of the five concentrations of each tested insecticides. Control diet cups were treated with 0.5 ml of acetone only. The lids of all diet cups had five holes for ventilation (3 mm diameter). Thirty 2-d-old adult of tarnished plant bug (mixed sex) were tested for each concentration of each insecticide.

Statistical Analysis

To calculate slopes and estimate lethal concentrations (LC₅₀) and resistance ratios (RR₅₀), mortality data for each bioassay (replication-generation) were analyzed with the PROBIT procedure, using log base 10 of the concentration. Χ ² goodness-of-fit tests were used to determine if response data followed the linear probit model for LC₅₀s (SAS Institute 2013). Differences in LC₅₀ values were considered significant if their 95% CL values did not overlap. Mortality for each bioassay was corrected for control effects using Abbott’s formula (Abbott 1925). Resistance ratios (RR₅₀) and confidence intervals were calculated using the method of Robertson and Priestler (1992). In order to estimate a real data variability, individual regressions that were not significant with the above criteria were not eliminated. PROC GLM procedure of SAS (SAS Institute 2013) was used to detect differences between bioassay treatments. Randomized complete blocks with a 2 × 11 factorial arrangement (bioassay treatments × replication) were used for LC₅₀, RR₅₀, mortality percentage, control mortality, mortality percentage at the highest concentration, percentage of significant regression, and percentage of positive dose-response. All these parameters were analyzed at each time period using one-way analysis of variance (ANOVA) followed by Tukey’s HSD. Variability among replicates was measured by LC₅₀s and RR₅₀s estimated values.

Results

Toxicity of Acephate by Glass-Vial Versus Diet-Overlay Assays

There were no statistically significant differences in toxicity of acephate to tarnished plant bug adults between the glass-vial and diet-overlay method at 6-h post-exposure, but there were significant differences at 24-, 48-, and 72-h post-exposure (Table 1). Although tarnished plant bug had a low sensitivity to this insecticide in diet, the probit model produced a good fit of the data among the concentrations tested in the bioassays to determine LC₅₀ values (Table 2). The lowest LC₅₀ value throughout all generations in the diet-overlay bioassay was 21.27 ± 18.0 μg per cup, while in the glass-vial assay it was 4.69 ± 18.0 μg per vial at 72-h post-exposure.

Table 1.

Comparison of glass-vial and diet cup bioassays for Lygus lineolaris (exposed to the insecticides acephate at different concentrations

Parameters	Insects	Hours AEX-Bioassay (means + SE)
		6-h		24-h		48-h		72-h
		Diet	Vial	Diet	Vial	Diet	Vial	Diet	Vial
LC50a	3,960	255.67 ± 80.63a	116.41 ± 32.32a	112.40 ± 6.52a	16.57 ± 2.23b	84.70 ± 21.90a	8.70 ± 1.04b	21.27 ± 8.00a	4.69 ± 0.51b
RR50	3,960	14.41 ± 3.62a	11.74 ± 6.49a	29.77 ± 8.05a	3.31 ± 1.41a	12.26 ± 3.69a	2.14 ± 0.72b	6.44 ± 1.20a	2.26 ± 0.19b
Dose-response (%)	3,960	54.55 ± 1.5a	90.94 ± 0.9a	90.91 ± 0.9a	100a	100a	100a	100a	100a
Significant regression (%)	3,960	0b	54.55 ± 1.5a	9.01 ± 0.9b	81.81 ± 1.20a	54.55 ± 1.50a	72.72 ± 1.40a	54.55 ± 0.15a	72.72 ± 1.40a
Mortality (M) control (%)	3,960	0b	4.08 ± 1.70a	0b	6.06 ± 1.73a	0.60 ± 0.04a	8.18 ± 2.12a	5.45 ± 0.81a	10.00 ± 2.69a
M highest concentration (%)	3,960	2.73 ± 8.39a	42.73 ± 0.87b	8.79 ± 1.81b	91.51 ± 2.43a	39.39 ± 3.35b	97.27 ± 1.78a	57.87 ± 4.36b	97.27 ± 1.78a

Means ± SE followed by the same letter in each row by hours are not significantly different (P < 005 Tukey’s test).

^aLethal concentration and resistance ratio values are in μg per vial or diet cup.

Table 2.

Lethal mortality response (LC₅₀) of Lygus lineolaris bioassayed using glass vial and diet cup with the insecticide acephate at different concentrations

Bioassay (B)-No.	Concentration response (μg per vial or diet cup)
	n	Slope ± SE	LC50 (95% CI)a	Probit trend				RR50 (95% CI)d
				Test for slopeb		Test for GoFc
				Χ 2	P > Χ 2	Χ 2	P > Χ 2
B1-Diet	180	0.26 ± 0.15	66.9 (-)	3.06	0.0804	0.23	0.8728	15.63 (4.28–63.64)
B2-Diet	180	0.37 ± 0.13	14.8 (5.79–61)	7.83	0.0052	0.38	0.7677	3.46 (0.53–8.88)
B3-Diet	180	0.28 ± 0.16	22.17 (-)	2.95	0.0861	0.16	0.9252	5.01(0.58–43.41)
B4-Diet	180	0.31 ± 0.15	56.21 (-)	4.07	0.0437	0.56	0.6407	13.14 (5.62–122.58)
B5-Diet	180	0.22 ± 0.14	106.4 (-)	2.22	0.1366	0.06	0.9829	24.85 (9.46–90.87)
B6-Diet	180	0.33 ± 0.08	101.07 (36.71–213.81)	16.52	<0.0001	2.54	0.0542	23.57(10.35–66.98)
B7-Diet	180	0.43 ± 0.15	4.28 (2.69–39.39)	7.84	0.0051	0.1	0.9597	1
B8-Diet	180	0.44 ± 0.34	10.79 (-)	1.64	0.1997	0.41	0.7500	2.55 (0.43–6.27)
B9-Diet	180	0.43 ± 0.09	82.96 (18.97–98.28)	22.20	<0.0001	2.35	0.0707	19.29 (10.38–24.34)
B10-Diet	180	0.24 ± 0.18	62.3 (-)	1.82	0.1771	0.11	0.9569	14.54 (0.32–80.01)
B11-Diet	180	0.33 ± 0.12	50.8 (9.51–95)	7.52	0.0061	0.001	0.9999	11.82 (3.55–21.34)
B1-Vial	180	0.93 ± 0.25	13.01(5.15–20.60)	14.13	0.0002	0.79	0.5001	6.93 (2.43–19.74)
B2-Vial	180	1.11 ± 0.44	7.74 (-)	6.21	0.0127	3.77	0.0101	5.39 (1.65–17.61)
B3-Vial	180	1.12 ± 0.28	14.02 (7.25–20.38)	15.60	<0.0001	1.33	0.2600	8.58 (3.4–21.67)
B4-Vial	180	2.42 ± 0.46	5.91(4.53–7.31)	27.18	<0.0001	0.47	0.7041	4.6 (2.06–10.25)
B5-Vial	180	1.55 ± 0.37	7.01(4.44–9.35)	17.57	<0.0001	0.41	0.7400	5.09 (2.19–11.83)
B6-Vial	180	0.53 ± 0.25	8.18 (-)	4.55	0.003	10.01	<0.0001	5.85 (0.24–5.53)
B7-Vial	180	6.24 ± 16,652.47	9.92 (-)	0	0.9997	1.53	0.2100	7.48 (1.66E-112–3.37E-113)
B8-Vial	180	0.72 ± 0.15	1.27 (0.41–2.35)	21.52	<0.0001	0.58	0.6300	1
B9-Vial	180	1.12 ± 0.29	9.23 (1.7–33.81)	15.00	0.0001	2.18	0.0877	7.13 (2.80–18.16)
B10-Vial	180	0.75 ± 0.15	10.42 (5.23–16.27)	26.40	<0.0001	0.41	0.7430	8.18 (3.23–20.74)
B11-Vial	180	0.94 ± 0.19	9.0 (5.02–13.45)	24.78	<0.001	0.98	0.4000	5.54 (2.21–13.91)

^aLC₅₀ and RR₅₀ values are in μg per vial or diet cup; mortality was scored at 48 h.

^bTest for slope significance indicates dose affects mortality.

^cTest for goodness-of-fit (GoF) significance indicates error from probit trend is greater than expected for simple binomial response.

^dRR₅₀ (resistance ratio) and 95% CI calculated using formula from Robertson and Priestler (1992). Differences among SR₅₀ values are significant if 95% CI do not include 1.0. RR₅₀ compare the LC₅₀s among the highest LC₅₀ as a control.

(-)Values for CI not calculated.

Lethal concentration and RR₅₀ values for both bioassays decreased significantly over the post-application period. However, the percentage mortality on the diet-overlay assay from 6- to 72-h post-exposure was too low to accurately estimate the LC₅₀ value. Significantly lower mortalities were observed in the control groups of the diet-overlay assay at all post-exposure time points (6-h: F_{(1, 21)} = 8.11, P ≤ 0.0104; 24-h: F_{(1, 21)} = 12.27, P ≤ 0.0057; 48-h: F_{(1, 21)} = 16.62, P ≤ 0.0022; and 72-h: F_{(1, 21)} = 3.38, P ≤ 0.0959), with the highest mortality (5.45 ± 0.81 [SE]) at 72-h post-exposure. The control mortality percentage in the glass-vial assay increased over time, from 4.08% ± 1.70 (SE) at 6-h post-exposure to 10.00 ± 0.81 (SE) 72-h post-exposure (Table 1).

Toxicity of Permethrin Determined by the Glass Vial and Diet Overlay

For permethrin, the probit model produced a good fit of the data to estimate LC₅₀ values for this insecticide, and there were no significant differences in LC₅₀ values between the vial and diet-overlay bioassays at any of the post-exposure time points: 24-h: F_{(1, 21)} = 0.15, P = 0.7025; 48-h (F_{(1, 21)} = 1.44, P = 0.2449; 72-h (F_{(1, 21)} = 1.47, P = 0.2397) post-exposure (Table 3).

Table 3.

Comparison of glass-vial and diet cup bioassays for Lygus lineolaris exposed to the insecticide permethrin at different concentrations

Parameters-insecticides	Insects	Hours AEX-Bioassay (means + SE)
		6-h		24-h		48-h		72-h
		Diet	Vial	Diet	Vial	Diet	Vial	Diet	Vial
LC50a	3,960	100.27 ± 18.03b	4.69 ± 0.51a	5.91 ± 0.51a	5.33 ± 1.41a	3.74 ± 0.47a	5.46 ± 1.35a	2.88 ± 0.31a	4.19 ± 1.03a
RR50	3,960	2.40 ± 0.17a	25.32 ± 16.45a	1.62 ± 0.17a	34.04 ± 25.13a	1.95 ± 0.29a	57.61 ± 41.95a	1.81 ± 0.28a	194.62 ±162.83a
Dose-response (%)	3,960	100a	100a	100a	100a	100a	100a	100a	100a
Significant regression (%)	3,960	72.72 ± 4.49a	54.55 ± 1.57a	72.72 ±1.41a	27.27 ±1.41b	100a	36.36 ±1.52b	81.81 ± 1.22a	54.55 ±1.57a
Mortality (M) control (%)	3,960	0b	7.27 ± 0.38a	0.30 ± 0.30b	9.09 ± 3.68a	2.42 ± 0.91b	13.03 ± 4.08a	4.24 ±1.10b	27.66 ± 9.16a
M highest concentration (%)	3,960	98.18 ± 0.93a	92.72 ± 4.18a	99.69 ± 0.30a	95.15 ± 3.37a	99.69 ±0.30a	99.09 ± 0.90a	99.69 ± 1.05a	99.09 ± 0.90a

Means ± SE followed by the same letter in each row by hours are not significantly different (P < 005 Tukey’s test).

^aLethal concentration and resistance ratio values are in μg per vial or diet cup.

There were no significant differences at any evaluation time for RR₅₀ values: 6-h: F_{(1, 21)} = 1.98, P = 0.1970; 24-h: F_{(1, 21)} = 1.68, P = 0.2313; 48-h: F_{(1, 21)} = 1.79, P = 0.2172; 72-h: F_{(1, 21)} = 1.41, P = 0.2733. However, the estimated RR₅₀ values for the glass-vial assay were much greater than those for the diet-overlay assay, indicating more variability occurred in estimates using the glass-vial bioassay (Table 3). Unlike the response to acephate, permethrin was found to be faster acting on an artificial diet with 100% dose-response mortality at all evaluation times and with no significant differences among replicates using different generations. Similar to acephate, control mortality percentage increased substantially within time in the glass-vial assay from 7.27 ± 0.38 (SE) 6-h post-exposure to 27.66 ± 9.16 (SE) 72-h post-exposure, while mortality for the controls was not >5% in the diet-overlay assay, even at 72-h post-exposure (Table 3). This strongly affected estimation of RR₅₀ values and the percentage of the significant regression in the glass-vial bioassay. Significance regressions for the diet-overlay bioassay ranged from 72 to 100% over time from 6- to 72-h post-exposure, respectively, while in the glass-vial assay the significance percentage fluctuated from 27 to 54% (Table 3). The lowest variability was observed 24-h post-exposure for both bioassays (Table 4).

Table 4.

Lethal mortality response (LC₅₀) of Lygus lineolaris bioassayed using glass vial and diet cup with the insecticide permethrin at different concentrations

Bioassay (B)-No.	Concentration response (μg per vial or diet cup)
	n	Slope ± SE	LC50 (95% CI)a	Probit trend				RR50 (95% CI)d
				Test for slopeb		Test for GoFc
				Χ 2	P > Χ 2	Χ 2	P > Χ 2
B1-Diet	180	0.83 ± 0.19	6.99 (2.90–10.97)	19.31	<0.0001	1.19	0.3120	2.35 (0.94–5.88)
B2-Diet	180	0.79 ± 0.28	7.45 (-)	8.02	0.0046	2.39	0.0664	2.50 (0.8–7.85)
B3-Diet	180	0.65 ± 0.13	3.90 (1.22–7.5)	25.20	<0.0001	0.69	0.5585	1.31(0.44–3.89)
B4-Diet	180	0.84 ± 0.33	6.57 (-)	6.59	0.0103	4.5	0.0037	2.21 (0.64–7.64)
B5-Diet	180	0.76 ± 0.15	5.88 (2.67–9.5)	26.44	<0.0001	1.5	0.2126	1.98 (0.79–0.4.96)
B6-Diet	180	1.19 ± 0.288	5.35 (2.94–7.38)	17.20	<0.0001	0.71	0.5430	1.37 (0.55–3.4)
B7-Diet	180	0.68 ± 0.13	2.97 (1.14–5.38)	28.45	<0.0001	0.85	0.4644	1
B8-Diet	180	0.92 ± 0.18	7.09 (3.8–10.46)	25.13	<0.0001	0.2	0.8937	2.38 (1.02–5.58)
B9-Diet	180	1.07 ± 0.31	8.84 (0.19–24.89)	12.32	0.0004	2.63	0.0500	2.84 (1.16–6.95)
B10-Diet	180	1.02 ± 0.39	4.9 (-)	6.82	0.009	4.03	0.0070	1.65 (0.55–4.92)
B11-Diet	180	0.84 ± 0.2	5.16 (2.03–8.25)	18.68	<0.0001	2.55	0.0539	1.73 (0.68–4.41)
B1-Vial	180	6.89 ± 16,280.9	11.03 (-)	0	0.9997	1.06	0.3653	257.89 (2.24E-104–2.96E-108)
B2-Vial	180	0.58 ± 0.42	9.06 (-)	1.89	0.1700	30.78	<0.0001	194.52 (0.02–181,689.87)
B3-Vial	180	6.17 ± 6,735.44	10.19 (-)	0	0	0	0	248.34 (4.57E-12–1.35)
B4-Vial	180	6.44 ± 13,161.97	10.34 (-)	0	0	0	0	251.64 (1.48E-46–4.27E-50)
B5-Vial	180	0.57 ± 0.1	2.95 (1.57–4.71)	36.03	<0.0001	0.79	0.4985	66.97 (0.0097–463,804.95)
B6-Vial	180	0.55 ± 0.25	3.01 (-)	4.84	0.0279	6.73	0.0002	73.72 (0.009–624,555.87)
B7-Vial	180	0.84 ± 0.41	0.98 (-)	4.21	0.0403	4.5	0.0037	24.01 (0.04–157.29)
B8-Vial	180	0.97 ± 0.37	0.74 (0.04–1.33)	6.84	0.0089	0.1	0.9574	7.84 (0.001–59,853.87)
B9-Vial	180	1.17 ± 0.28	9.79 (5.57–13.77)	18.07	<0.0001	0.96	0.4112	206.08 (0.03–1,419,632.49)
B10-Vial	180	0.32 ± 0.17	0.04 (-)	3.72	0.0538	0.31	0.8152	1
B11-Vial	180	0.44 ± 0.3	0.60 (-)	2.17	0.14	8.35	<0.0001	9.59 (0.0007–14,1051.06)

^aLC₅₀ and RR₅₀ values are in μg per vial or diet cup; mortality was scored at 24 h.

^bTest for slope significance indicates dose affects mortality.

^cTest for goodness-of-fit (GoF) significance indicates error from probit trend is greater than expected for simple binomial response.

^dRR₅₀ (resistance ratio) and 95% CI calculated using formula from Robertson and Priestler (1992). Differences among SR₅₀ values are significant if 95% CI do not include 1.0. RR₅₀ compare the LC₅₀s among the highest LC₅₀ as a control.

(-)Values for CI not calculated.

Toxicity of Thiamethoxam Determined by the Glass Vial and Diet Overlay

Similar to permethrin, thiamethoxam was very toxic to tarnished plant bug adults (Table 5). There were statistically significant differences in LC₅₀ values for thiamethoxam to tarnished plant bug adults between bioassays at all evaluation times, 6-h: F_{(1, 21)} = 1.88, P = 0.1861; 24-h: F_{(1, 21)} = 14.32, P = 0.0012; 48-h: F_{(1, 21)} = 3.58, P = 0.0731, except for 72-h (F_{(1, 21)} = 22.44, P = 0.0001) post-exposure. No significant differences were observed at any evaluation time for RR₅₀. The RR₅₀ estimated values in the diet-overlay assay ranged from 1.03-fold for generation five (F₅) to 4.23-fold for the next generation 6 (F₆). For the glass-vial assay, RR₅₀ estimates ranged from 2.32-fold to 1,823-fold for the second (F₂) and 10th generations (F₁₀), respectively (Table 5). Similar to permethrin, the probit model produced a good fit for the data among the bioassays used in the analyses used to determined LC₅₀ values for this insecticide, with no significant differences in the percentage mortality at 6, 24, 48, or 72 h (100% mortality at all time points) for either bioassays method (Table 4). The LC₅₀ values exponentially decreased over the test period mainly on the diet-overlay bioassay. Significantly lower mortality in the control was observed in the diet-overlay assay at all evaluation times, with the highest control mortality of 1% ± 0.64 (SE) at 72-h post-exposure. Control mortality in the glass-vial assay was 4.85% ± 1.37 (SE) at 6-h post-exposure, increasing to 13.93 ± 3.26 (SE) at 72-h post-exposure (Table 6).

Table 5.

Comparison of glass-vial and diet cup bioassays for Lygus lineolaris exposed to the insecticides thiamethoxam at different concentrations

Parameters-insecticides	Insects	Hours AEX-Bioassay (means + SE)
		6-h		24-h		48-h		72-h
		Diet	Vial	Diet	Vial	Diet	Vial	Diet	Vial
LC50a	3,960	106.06 ± 25.22a	6.80 ± 5.09b	33.99 ± 4.89a	6.41 ± 5.40b	7.51 ± 3.62a	5.74 ± 5.06a	2.11 ± 1.09a	0.889 ± 0.42b
RR50	3,960	4.42 ± 1.17a	110,78.69 ± 9,834.59a	2.63 ± 0.29a	1,053.55 ± 684.11a	2.37 ± 0.85a	353.39 ±183.14a	4.51 ± 1.21a	744.78 ± 598.11a
Dose-response (%)	3,960	100a	100a	100a	100a	100a	100a	100a	100a
Significant regression (%)	3,960	18.18 ± 1.22a	45.45 ± 1.57a	54.55 ±1.57a	54.55 ± 1.57a	90.90 ± 0.90a	81.81 ± 1.21a	81.82 ± 1.23a	63.64 ± 1.52a
Mortality (M) control (%)	3,960	0b	4.85 ±1.376a	0b	7.00 ± 2.19a	0.33 ± 0.33b	8.48 ± 2.39a	0.90 ± 0.64b	13.93 ± 3.26a
M highest concentration (%)	3,960	24.10 ± 4.99b	72.42 ± 6.08a	53.94 ± 4.51b	82.00 ± 4.53a	82.72 ± 4.77a	94.85 ± 2.99a	84.55 ± 3.54b	98.18 ± 1.81a

Means ± SE followed by the same letter in each row by hours are not significantly different (P < 005 Tukey’s test).

^aLethal concentration and resistance ratio values are in μg per vial or diet cup.

Table 6.

Lethal mortality response (LC₅₀) of Lygus lineolaris bioassayed using glass vial and diet cup with the insecticide thiamethoxam at different concentrations

Bioassay (B)-No.	Concentration response (μg per vial or diet cup)
	n	Slope ± SE	LC50 (95% CI)a	Probit trend				RR50 (95% CI)d
				Test for slopeb		Test for GoFc
				Χ 2	P > Χ 2	Χ 2	P > Χ 2
B1-Diet	180	0.57 ± 0.16	2.38 (0.57–24.37)	13.23	0.0003	4.04	0.007	1.14 (0.99–3.31)
B2-Diet	180	0.48 ± 0.10	8.44 (3.97–15.56)	20.94	<0.0001	0.81	0.487	4.04 (0.74–4.56)
B3-Diet	180	0.45 ± 0.11	8.27 (2.58–17.18)	17.54	<0.0001	1.86	0.1344	3.99 (0.71–5.82)
B4-Diet	180	0.46 ± 0.13	3.95 (1.64–24.29)	12.65	0.0004	0.44	0.7234	1.91 (0.93–19.74)
B5-Diet	180	0.58 ± 0.18	2.15 (0.64–95.5)	10.45	0.0012	5.35	0.0011	1.03 (0.91–4.78)
B6-Diet	180	0.52 ± 0.12	8.54 (0.24–13.35)	17.52	<0.0001	1.19	0.3105	4.23 (2.22–3.77)
B7-Diet	180	0.52 ± 0.08	2.07 (0.97–7.47)	39.76	<0.0001	2.02	0.1089	1
B8-Diet	180	0.47 ± 0.17	4.99 (-)	7.57	0.0059	4.58	0.0033	2.44 (0.34–8.12)
B9-Diet	180	0.46 ± 0.09	3.25 (0.44–34.71)	28.32	<0.0001	2.61	0.0494	1.57 (0.06–4.51)
B10-Diet	180	0.82 ± 0.24	4.5 (2.21–10.3)	12.01	0.0005	0.7	0.5500	1.96 (0.68–4.06)
B11-Diet	180	1.25 ± 0.36	2.19 (0.36–7.26)	12.24	0.0005	4.52	0.0036	1.07 (0.31–6.74)
B1-Vial	180	0.24 ± 0.10	0.02(4.99E-14–0.22)	5.43	0.0197	0.25	0.8631	1
B2-Vial	180	0.31 ± 0.12	0.04 (3.23E-7–0.28)	6.91	0.0086	0.78	0.5069	2.32 (0.004–1,423.49)
B3-Vial	180	1.22 ± 0.30	0.59 (0.35–0.81)	16.65	<0.0001	0.11	0.9600	67.55 (0.38–12,080.12)
B4-Vial	180	0.64 ± 0.27	3.39 (-)	5.53	0.0187	5.92	0.0005	334.43 (1.56–71,628.72)
B5-Vial	180	0.46 ± 0.15	0.1 (0.0008–0.32)	9.57	0.0020	1.09	0.3531	10.88 (0.04–2,719.04)
B6-Vial	180	0.34 ± 0.13	0.03 (9.77E-8–0.22)	6.63	0.0100	1.19	0.3100	3.65 (0.008–1,707.55)
B7-Vial	180	4.55 ± 1.82	0.75 (0.6–2.8)	6.25	0.0124	0.67	0.5694	92.23 (0.52–16,442.27)
B8-Vial	180	0.3 ± 0.09	0.21 (0.005–0.7)	11.97	0.0005	1.07	0.3584	18.93 (0.08–4,751.41)
B9-Vial	180	0.54 ± 0.15	1.6 (0.0005–11.78)	13.87	0.0002	2.98	0.0300	195.99 (0.94–40,711.31)
B10-Vial	180	0.35 ± 0.34	56.23 (-)	1.07	0.3001	2.62	0.0488	1,823.77 (4.56–733,498.33)
B11-Vial	180	0.78 ± 0.31	0.19 (0.001–0.41)	6.48	0.0100	0.17	0.9200	20.84 (0.10–4,362.07)

^aLC₅₀ and RR₅₀ values are in μg per vial or diet cup; mortality was scored at 48 h.

^bTest for slope significance indicates dose affects mortality.

^cTest for goodness-of-fit (GoF) significance indicates error from probit trend is greater than expected for simple binomial response.

^dRR₅₀ (resistance ratio) and 95% CI calculated using formula fromRobertson and Priestler (1992). Differences among SR₅₀ values are significant if 95% CI do not include 1.0. RR₅₀ compare the LC₅₀s among the highest LC₅₀ as a control.

(-)Values for CI not calculated.

Toxicity of Sulfoxaflor Determined by the Glass Vial and Diet Overlay

The LC₅₀ values for sulfoxaflor for both bioassays were generally higher than those for thiamethoxam or permethrin, but lower than those for acephate. There were no significant differences between the two bioassays in LC₅₀ estimates at any evaluation time (Table 7), although adults tested using the glass-vial method had a better fit of the data than did those tested with the diet-overlay method with higher percent mortality at 6- and 24-h post-exposure. Similarly, for the RR₅₀ values, there were no significant differences at any time point; however, the RR₅₀ values for the glass-vial assay were higher than those for the diet-overlay assay; yet showing a higher variability on the glass-vial bioassay (Table 8). Similar to acephate, sulfoxaflor was slow-acting when administered on either the artificial diet or via the glass-vial assay compared with permethrin and thiamethoxam. Control mortality increased over time in the glass-vial assay, from 6.6% at 6-h to 25.5% at 72-h post-exposure, while control mortality in the diet-overlay assay was 0% at 6-h and 5.8% at 72-h post-exposure (Table 8). The lowest variability (the condition with the most consistent response) was observed 72-h post-exposure in both bioassays (Table 8).

Table 7.

Comparison of glass-vial and diet cup bioassays for Lygus lineolaris exposed to the insecticides sulfoxaflor at different concentrations

Parameters-insecticides	Insects	Hours AEX-Bioassay (means + SE)
		6-h		24-h		48-h		72-h
		Diet	Vial	Diet	Vial	Diet	Vial	Diet	Vial
LC50a	3,960	361.80 ± 79.69a	90.96 ± 39.00b	143.05 ± 35.32a	33.16 ± 15.01b	39.09 ± 39.01a	6.91 ± 2.35b	6.13 ± 3.55a	2.89 ± 0.64b
RR50	3,960	2.63E+06 ± 2.64+E05a	2.10E+06 ± 1.26E+05a	123.93 ± 50.22a	127.23 ± 45.79a	9.35 ± 2.43a	20.19 ± 9.61a	2.57 ± 0.71a	58.46 ± 53.12a
Dose-response (%)	3,960	45.45 ± 1.57a	90.90 ± 0.90a	81.81 ± 12.19a	100a	100a	100a	100a	100a
Significant regression (%)	3,960	0b	63.63 ± 1.52a	27.27 ± 1.40b	63.63 ± 1.52a	72.72 ± 1.04a	63.63 ± 1.52a	90.90 ± 0.90a	63.63 ± 1.52a
Mortality (M) control (%)	3,960	0b	6.6 ± 1.64a	0b	9.09 ± 1.35a	1.21 ± 0.50b	13.64 ± 2.96a	5.75 ± 1.35b	25.45 ± 7.60a
M highest concentration (%)	3,960	3.64 ± 1.58b	48.67 ± 7.98a	18.67 ± 5.33b	73.63 ± 6.64a	52.42 ± 5.66b	88.79 ± 3.52a	80.30 ± 3.28a	96.36 ± 1.22a

Means ± SE followed by the same letter in each row by hours are not significantly different (P < 005 Tukey’s test).

^aLethal concentration and resistance ratio values are in μg per vial or diet cup.

Table 8.

Lethal mortality response (LC₅₀) of Lygus lineolaris bioassayed using glass vial and diet cup with the insecticide sulfoxaflor at different concentrations

Bioassay (B)-No.	Concentration response (μg per vial or diet cup)
	n	Slope ± SE	LC50 (95% CI)a	Probit trend				RR50 (95% CI)d
				Test for slopeb		Test for GoFc
				Χ 2	P > Χ 2	Χ 2	P > Χ 2
B1-Diet	180	1.85 ± 0.53	3.20 (09.01–4.05)	12.15	0.0005	0.71	0.5418	1.38 (0.91–7.26)
B2-Diet	180	1.59 ± 0.56	7.85 (2.08–4.75)	7.97	0.0048	2.35	0.0696	3.39 (1.33–6.79)
B3-Diet	180	0.56 ± 0.15	7.94 (2.08–15.36)	12.45	0.0004	0.50	0.6786	3.43 (0.31–6.79)
B4-Diet	180	0.88 ± 0.22	8.39 (1.91–28.01)	15.87	<0.0001	0.69	0.5538	3.61 (1.58–7.34)
B5-Diet	180	0.30 ± 0.74	9.09 (3.79–25.44)	16.49	<0.0001	0.47	0.6993	3.95 (0.58– 4.71)
B6-Diet	180	0.66 ± 0.19	2.37 (0.05–14.01)	11.37	0.0007	0.65	0.5806	1.06 (0.89–9.63)
B7-Diet	180	0.42 ± 0.08	2.31 (0.63–9.91)	26.68	<0.0001	0.30	0.8239	1
B8-Diet	180	0.76 ± 0.29	8.25 (-)	6.79	0.0092	2.43	0.0629	3.57 (0.63–5.56)
B9-Diet	180	0.61 ± 0.13	2.52 (6.76–20.23)	21.24	<0.0001	0.53	0.8588	1.07 (0.05–5.35)
B10-Diet	180	0.50 ± 0.17	4.32 (21.77–524.60)	8.17	0.0043	0.28	0.8361	1.89 (0.83–11.01)
B11-Diet	180	0.47 ± 0.10	8.52 (2.23–17.36)	15.23	<0.0001	0.99	0.3921	3.70 (0.62–5.24)
B1-Vial	180	0.53 ± 0.31	3.13 (-)	2.86	0.0911	4.53	0.0035	143.91 (0.44–46,766.95)
B2-Vial	180	2.05 ± 3.46	3.73 (-)	0.52	0.4699	0.25	0.8591	288.56 (1.50–55,532.61)
B3-Vial	180	0.20 ± 0.08	0.02 (2.13E-13–0.25)	5.55	0.0185	0.28	0.8398	1
B4-Vial	180	0.89 ± 0.22	6.17 (2.75–9.48)	16.37	<0.0001	0.92	0.4282	484.12 (2.98–78,518.96)
B5-Vial	180	1.10 ± 0.20	1.28 (0.84–1.79)	30.23	<0.0001	0.88	0.4458	109.64 (0.69–17,423.16)
B6-Vial	180	0.73 ± 0.13	6.95 (3.99–10.83)	30.35	<0.0001	1.17	0.3192	536.27 (3.33–86,183.4)
B7-Vial	180	0.74 ± 0.41	2.51 (-)	3.19	0.0739	7.86	<0.0001	191.47 (0.88–41,495.32)
B8-Vial	180	0.91 ± 0.68	1.85 (-)	1.78	0.1819	5.63	0.0007	95.53 (0.32–28,784.28)
B9-Vial	180	0.72 ± 0.11	1.47 (0.89–2.18)	38.8	<0.0001	0.22	0.8780	132.25 (0.83–21,130.05)
B10-Vial	180	0.51 ± 0.12	1.12 (0.23–2.57)	17.62	<0.0001	0.80	0.4929	24.68 (0.12–4,897.20)
B11-Vial	180	0.57 ± 0.16	3.50 (0.00–20.30)	12.23	0.0005	2.55	0.0536	211.87 (1.19–37,563.79)

^aLC₅₀ and RR₅₀ values are in μg per vial or diet cup; mortality was scored at 72 h.

^bTest for slope significance indicates dose affects mortality.

^cTest for goodness-of-fit (GoF) significance indicates error from probit trend is greater than expected for simple binomial response.

^dRR₅₀ (resistance ratio) and 95% CI calculated using formula from Robertson and Priestler (1992). Differences amongst SR₅₀ values are significant if 95% CI do not include 1.0. RR₅₀ compare the LC₅₀s among the highest LC₅₀ as a control.

(-)Values for CI not calculated.

Discussion

Tarnished plant bug is the most important pest of cotton in the Mississippi Delta (Catchot 2020) and it has traditionally been managed using several practices, such as field selection and planting arrangements, variety selection, and manipulation of planting date. However, the most important control method is chemical control (Snodgrass and Gore 2007, Gore et al. 2014). Surveys of pesticide resistance among tarnished plant bug populations can provide an early warning of the need to modify Integrated Pest Management (IPM) strategies (Miller et al. 2010). Laboratory pesticide bioassays are a convenient and useful way to screen insecticides for efficacy and to choose effective materials for control. Historically, Lygus spp. were examined for pesticide resistance in the United States as early as 1954 (Menke 1954), when several species of Lygus were found to be resistant to dichlorodiphenyltrichloroethane (DDT) (Portilla et al. 2018). Andres et al. (1955) and Leigh and Jackson (1968) conducted the first experiments with the glass-vial method; since then, many other studies have used the glass-vial essay method with or without modifications, together with dose-response regressions. This method became the principal method for the annual ARS-USDA insecticide resistance survey in the Mississippi Delta, starting in 1994 when tarnished plant bug was found to be resistant to pyrethroids (Portilla et al. 2018).

In the present study, observations were limited to experiments against a single laboratory colony throughout 11 generations to ensure homogeneity (specific characteristics and similar age) among the replicates (= the generations) and treatments (pesticides × assay method). This study demonstrated that the levels of resistance of this colony could have a wide variation in its response to the pesticides examined, either via glass-vial or diet-overlay bioassays. However, there was greater variation in the level of pesticide resistance in the glass-vial bioassay than in the diet-overlay assay, except for acephate. The variation in our acephate bioassay (values ranged from 1.27 to 14.02 μg per vial) was comparable to that found by Parys et al. (2017), who obtained even higher variability with LC₅₀ values ranging from 6.16 to 75.62 μg per vial. Portilla et al. (2018) reported a mean LC₅₀ value for the same laboratory colony of 17.1 μg per vial of acephate, which increased 52.3 μg per vial when used on floral-foam at 24-h post-exposure. This could explain why in the present study tarnished plant bug had a much lower response to acephate in the diet-overlay bioassay (4.28–106.4 μg per diet cup) (Table 1). Although acephate is considered an insecticide with very low adsorption (K_oc 2.7) (Yen et al. 2000), it is one of the few insecticides that is highly soluble in acetone as well as water (>1,000 ppm) (Wang et al. 2007). Since the diet is a 50% water-based solution, this insecticide will reach the bottom of the diet cup (19-mm depth) more quickly than the other insecticides like permethrin that has <10 ppm solubility in water (SPEX CertiPreps’s Guide to Pesticide Solubility 2017). This highwater solubility may have caused acephate have to leach downward, leaving only a low residual level of the insecticide on the surface of the diet, while permethrin’s residues on the diet surface would have remained higher. The relative mobility of the insecticide in the diet may also have moderately affected the LC₅₀ for sulfoxaflor as it is shown in Table 7. M. Portilla et al. (unpublished data) observed that the LC₅₀ value could be lowered in a diet-overlay bioassay by reducing the amount of diet (to 2.0 ml per cup, with a 5-mm depth), resulting in LC₅₀ values for acephate that were about 30% of that in the present study (LC₅₀ of 3.9–35 μg per diet cup), but no such effect was observed for permethrin, thiamethoxam, or sulfoxaflor. Although the LC₅₀ values for acephate of M. Portilla (unpublished data) using the diet-overlay (2 ml diet per cup) method could be considered a little higher than those obtained in the glass-vial assay, the lower variability among replicates in the diet bioassay suggest the diet bioassay may be more suitable to separate susceptible and resistant tarnished plant bug field populations (M. Portilla, unpublished data).

The adsorption of the insecticides by diet in the diet-overlay assay could be comparable under normal circumstances to the inner surface of soil and/or plants in the field, delaying the knockdown effects that normally are found when compounds’ LC₅₀ values measured using the glass-vial method. Therefore, the LC₅₀s from a glass-vial assay may be inaccurate, although it has been adequate for monitoring purposes with fast-acting insecticides such as permethrin. Based on the results of our tests, permethrin and thiamethoxam were found to cause more rapid mortality with the glass-vial method compared to the diet-overlay assay. The LC₅₀ values from the glass-vial assay did not differ between the 6- and 24-h post-exposure for either permethrin or thiamethoxam; whereas in the diet-overlay assay, the highest level of morality was not observed until the 24-h time point for permethrin and 48-h for thiamethoxam. The early mortality observed at 6-h post-exposure for thiamethoxam in the glass-vial assay with pesticide solution presented in floral-foam was unexpected because the earliest knockdown effect reported for tarnished plant bug was 24-h post-exposure (Snodgrass et al. 2008b). Yet the LC₅₀ found for thiamethoxam either at 24-h post-exposure in the glass-vial assay with floral-foam and 48-h post-exposure in the diet-overlay assay was too high if compared with the 24-h LC₅₀s reported by Snodgrass et al. (2008b), which ranged from 0.35 to 2.30 at a concentration of 15 μg per vial in 2006 and from 0.64 to 2.18 at a concentration of 15 μg per vial in 2007 from tarnished plant bug field populations collected in the Mississippi Delta. However, both bioassays reached those values 72-h post-exposure with 0.89 for 15 μg per vial in the glass-vial assay and 2.11 for 15 μg per cup in the diet-overlay assay, suggesting either that this colony has some levels of tolerance to thiamethoxam or the field colony is not very resistant to this insecticide. Either way, direct comparison between the results of this study and projected field resistance is not possible; at least, the comparison would be between our colony and possible wild resistant populations reared under laboratory conditions and fed with artificial diet.

The quick knockdown effect of permethrin in both bioassays might be particularly important, since it is one of the most commonly used insecticides in cotton for control of tarnished plant bug. Snodgrass and Scott (1999) found that permethrin could cause mortality as early as 3-h post-exposure at a treatment level of 15 μg per vial. They reported 95% and 98% mortality 3-h post-exposure in susceptible laboratory and field tarnished plant bug population, while three resistant field populations reared under laboratory conditions reported 30%, 48%, and 78% mortality, respectively. They also mentioned that the mortality of the same populations varied if dosage per vial was changed to 10 or 20 μg, resulting in low mortality in susceptible insects 3-h post-exposure and high mortality for resistance insects 1-h post-exposure, respectively. Based on our results, the LC₅₀ values for permethrin were 5.33 μg in the glass-vial assay and 5.91 μg in the diet cup bioassay, both at 24-h post-exposure, and these values did not differ greatly from those reported by Snodgrass (1996a) (1.0-fold higher, i.e., 4.6 μg per vial at 24-h post-exposure) or of Snodgrass and Scott (1999) (1.5-fold higher, i.e., 3.4 μg per vial at 24-h post-exposure) using susceptible tarnished plant bug populations reared under laboratory conditions. This suggests that our laboratory colony could be considered susceptible to permethrin if compared with the LC₅₀s reported by Snodgrass and Scott (1999) (values ranging from 11.1 to 62.0 μg per vial 24-h post-exposure) and Snodgrass et al. (2009) (values ranging from 29.0 to 62.2 μg per vial 24-h post-exposure) on resistant tarnished plant bug collected from the field. However, it is important to mention that Snodgrass (1996b) and Parys et al. (2018) found higher LC₅₀s for laboratory colonies than that for susceptible field populations (1.2-fold lower than that of the laboratory colonies). This difference in LC₅₀ values between the susceptible laboratory colony and field populations may have been caused by a variety of factors, including greater handling mortality for field populations, unknown age of field bugs, stress from change in diet after collection, or heat stress. None of these factors would have affected the laboratory colony, making susceptible field populations even less tolerant than normal.

There is little information on baseline assay data concerning tarnished plant bug resistant to sulfoxaflor (Portilla et al. 2018). Parys et al. (2017) reported LC₅₀ values for tarnished plant bug laboratory colonies ranging from 2.59 to 27.18 μg per vial, while those from field populations from the Mississippi Delta ranged from 0.25 to 45.81 μg per vial in a 24-h glass-vial assay. However, it is unknown if the assayed individuals came from a susceptible population or from a field population that had already evolved resistance to sulfoxaflor. On the other hand, the LC₅₀ value for sulfoxaflor of 27.18 μg per vial 24-h post-exposure for the laboratory colony indicated that these insects could be considered moderately resistant via the commonly used glass-vial bioassay, despite the fact that the colony has not been exposed to this insecticide. Nor is there any reason to believe that the bugs in this colony are fitter than field insects and thus requiring higher doses of insecticide to be killed. Regardless, these results differed from those of our investigation where 33.16 μg per vial and 143.3 μg per cup were the LC₅₀ values in the glass-vial and diet-overlay bioassays at 24-h post-exposure, respectively. Our results also differed from those of Portilla et al. (2018) who tested our laboratory colony using the glass-vial method and reported 15.9 μg per vial and 53.2 μg per vial when pesticide was presented on floral-foam at 24-h post-exposure. All these results indicate high variability regardless of the bioassay method used, suggesting that this insecticide needs further investigation for future resistance-monitoring procedures.

In general, these results demonstrate, via the measured level of mortality from permethrin or thiamethoxam through contact or feeding-based bioassay, and of acephate through the glass-vial bioassay, the potential of these insecticides to control tarnished plant bug. The rapid induction of mortality found in the glass-vial method explains why this bioassay has been considered for decades to be one of the quickest and most efficient methods for identifying insecticide resistance. The wide variation in tarnished plant bug mortality response could be a limitation in accuracy as was clearly observed in this study when a colony with similar characteristics was exposed to the insecticides. Miller et al. (2010) noted that rapid assay methods are important for early detection of impending pesticide resistance, but more accurate data are needed to improve chemical control use strategies. As mentioned before, insecticide adsorption in the diet could be comparable to what happen in the field under normal circumstances, making this diet-overlay system an appropriate assay for separating susceptible and resistant populations with small range of mortality responses under field conditions. Zhang et al. (2015) mentioned that the accuracy and usefulness of bioassays would depend on the correlation of bioassay results with methods that more closely approximate the effects of insecticides used under field conditions.